Hi everyone, this is Mr. Daniel. How are you?

I hope you guys are having a wonderful Easter holiday weekend and enjoying the wonderful southern California Spring! I would like to share with you the use of LiDAR and CAMERA on Self Driving Electric Vehicles (EV)

LiDAR stands for Light Detection and Ranging and uses lasers to detect objects. LiDAR uses light pulses to bounce off objects millions of times per second. This generates a 3D representation of the world allowing the EV to “see” and make decisions used in autonomous driving.

The typical wavelength used for automotive lidar is around 900nm or 1500nm. This is in the Infra-red region of the EM Spectrum and is beyond the wavelength range that our eyes can perceive (400nm to 750nm) and therefore we are not able to see Lidars shooting out of the top of cars.

LiDAR versus Camera, Other EV manufacturers versus Tesla. The arms race to reach full autonomous driving. LiDAR vs Camera has been an ongoing debate amongst car manufacturers. Waymo, the autonomous car segment owned by Alphabet, firmly believes in the use of LiDar technology as the way to real autonomous driving. (Alphabet is google’s parent company, it was formed in 2015 when google renamed itself to Alphabet and made Google its subsidiary to avoid antitrust violations). Tesla, on the other hand, decided to approach autonomy with Camera, with help of radar and Artificial Intelligence. Tesla uses Cameras (located on a vast fleet of Tesla Model S, 3, X, Y’s on the road) to capture the 3D world and use machine learning to “see” the world.

“They’re all going to dump lidar,” Elon Musk said at an April event showcasing Tesla’s self-driving technology. “Anyone relying on lidar is doomed.” “Lidar is a fool’s errand,” Elon Musk said. “Anyone relying on lidar is doomed. Doomed! [They are] expensive sensors that are unnecessary. It’s like having a whole bunch of expensive appendices. Like, one appendix is bad, well now you have a whole bunch of them, it’s ridiculous, you’ll see.”

Waymo CEO argued that Tesla’s current strategy is unlikely ever to produce a fully self-driving system and commented “This is not how it works”

Austin Russel, 26 years old Billionaire after his company LUMINAR went on Public in 2020 said,

“What Tesla has today, what they call full self-driving you don’t need lidar for. The problem is that it’s not fully self-driving (laughs), it’s actually not self-driving at all. That’s where I think they have drawn huge criticism in the industry with this rogue branding approach to what is being delivered”

So what is going on? We have different theories on driverless technology!

Here are some advantages of LiDAR

  1. It is extremely accurate and precise. It can detect what direction a pedestrian is facing as well as hand movements from a distance.
  2. LiDAR can generate an extremely accurate 3D representation of the real world compared to Cameras. The argument is the LiDAR will not be tricked by shadows that Cameras are susceptible to
  3. LiDAR can calculate distances, and map topography accurately with fast speed and reasonable cost.
  4. LiDAR does not depend on ambient light (visible light) since it uses Infrared and is less susceptible to rain and fog. Cameras will struggle in bad weather conditions like heavy rain and fog. Tesla reasons that an easy fix is to use simple radar to supplement cameras in bad lighting conditions.

Here are some disadvantages of LiDAR

  1. LiDAR is much more costly than Cameras. This is one of the reasons for Tesla’s choice of Camera systems in their EV.
  2. LiDAR can not interpret words (traffic signs), colors (light traffic light). Cameras, along with AI and complex neural network computing, can solve these issues. Cameras are able to solve the vision problem. Tesla argues that visual recognition via the Camera is necessary for Autonomy.
  3. LiDAR can only process spatial information (object and distances) but not the complexities we have in real-world driving situations. For example, LiDAR would have trouble distinguishing a trash bag vs a rock on the freeway.
  4. LiDAR is UGLY, have you seen it on top of cars, check out any Waymo’s fleet. It looks like a giant pimple on a car.

What do you guys think? I am on team Camera. Machine learning using billions of miles of camera feeds and complex neural computing is the closest representation of the human visual and driving situations. There are situations that require interpretation instead of simple spatial representation. AI and Machine learning would be able to solve the visual problem and I doubt LiDAR alone can achieve autonomy.

So if you are team LiDAR, check out this video.

TEAM LiDAR: Watch this Video. It has great information! https://youtu.be/ibIzthKKZWY

And if you are team Tesla, check out this video

TEAM Camera: Watch this Video https://youtu.be/HM23sjhtk4Q

Almost everyone is using LiDAR. Waymo, Nio, Uber, Ford, Toyota, but not Tesla. Check out what Elon Musk has to say in the Team Tesla video and decide for yourself.

Tesla will release its FSD Beta 9.0 very soon. (FSD = Full Self Driving). You can watch how Tesla cars drive on their own in crowded city situations (with driver supervision at the moment).

In either case, regardless of which team you are on, hybrid use of the two technologies may be the most sensible solution to level 5 autonomy. Until then, game on!

Stay safe and take care,

Mr. Daniel

 

大家好，我是丹尼爾。

希望你們度過了一個愉快的假期週末，並享受著美好的南加州春天！ 我想與您分享LiDAR光學雷達和CAMERA攝像鏡頭在自動駕駛電動汽車（EV）上的使用LiDAR（光學雷達）是一種光學和測距技術，LiDAR是使用脈衝激光每秒將物體反彈數百萬次來檢測目標的距離。LiDAR可測量數千個點，以構建出感測器周圍環境的細節和精確的3D視野，從而使EV能夠“看到”並做出自動駕駛中使用的決策。

用於汽車光學雷達的典型波長約為900nm或1500nm。 它位於EM光譜的紅外區域，並且超出了我們的眼睛可以感知的波長范圍（400nm至750nm），因此我們看不到激光雷達從汽車頂部射出。

其他電動汽車製造商與特斯拉一直在使用光學雷達與攝像頭在全自動駕駛之間， 各有不同見解而正在進行辯論。   Alphabet擁有的自動駕駛汽車細分市場Waymo堅信，使用LiDar技術才是作為實現真正自動駕駛的方式。  （Alphabet是Google的母公司，成立於2015年，當時Google重命名為Alphabet，並成為Google的子公司以避免違反反托拉斯法）。 另一方面，特斯拉決定在雷達和人工智能的幫助下透過Camera攝像頭來實現自主。 特斯拉使用攝像頭（位於路上的特斯拉Model S，3，X，Y龐大的車隊中）捕獲3D世界並使用機器學習來“看到”世界。

埃隆·馬斯克（Elon Musk）在四月份展示特斯拉自動駕駛技術的活動上說 “他們都會拋棄光學雷達，”任何依賴光學雷達注定要失敗！[它們]這些昂貴的傳感器不是必要的，“ 比如說，一個附加物已經很糟糕，那麼現在您有一大堆，這是很荒謬的。”

Waymo首席執行官辯稱，特斯拉當前的戰略不太可能生產出完全自動駕駛的系統，並評論說“這不是它的工作原理”

26歲的億萬富翁奧斯汀·羅素（Austin Russel）在其公司LUMINAR於2020年上市後表示，

“特斯拉今天擁有什麼，他們所謂的全自動駕駛就是您不需要光學雷達。 但是問題在於它不是完全自動駕駛（笑），而實際上它根本不是自動駕駛，這就是我認為他們這種對所交付產品進行的方法態度使得在市場上廣泛被批評的地方。”

那麼發生了什麼？ 關於無人駕駛技術，我們有不同的理論！

以下是LiDAR的一些優勢

  1. 這是非常準確和精確的。它可以檢測行人面對的方向以及遠處手部的動作。
  2. 與相機相比，LiDAR可以以極其精確的3D顯示生在真實世界。重點是光學雷達不會像相機，容易受到的陰影所欺騙
  3. LiDAR可以計算距離，並以快速，合理的成本準確地繪製地形圖。
  4. LiDAR不依賴於環境光（可見光），因為它使用紅外光，並且不易受到雨水和霧氣的影響。相機會在惡劣的天氣條件下（例如大雨和霧就大幅降低了準確度。特斯拉（Tesla）認為，簡單的解決方法是在惡劣的照明條件下使用簡單的雷達來補充攝像頭。

LiDAR光達的一些缺點

  1. 光學雷達比攝像頭昂貴得多。這是特斯拉在電動汽車中選擇攝像系統的原因之一。
  2. LiDAR無法解釋辨識字詞（交通標誌），顏色（交通信號燈）。相機結合AI和複雜的神經網絡計算則可以解決這些問題，相機能夠解決視覺問題。特斯拉認為，通過相機進行視覺識別對於自主性是必要的。
  3. LiDAR僅能處理空間信息（物體和距離），而不能處理現實駕駛情況下的複雜性。例如，光學雷達很難區分垃圾袋和高速公路上的石頭。
  4. 光學雷達外觀不美觀。您是否在汽車上看到過光學雷達，請查看Waymo的所有機種，它外型看起來像是汽車上的巨型丘疹。

你們有什麼想法嗎？ 我是偏向支持Camera的一群，

透過數十億英里遠距離攝像的反覆測試，如同進行複雜的類似神經網絡計算.促使機器學習達到最接近人類視覺和駕駛情況。 有些狀況不僅僅是簡單的空間呈現而是需要多的詮釋。人工智能和機器學習系統將能夠解決視覺問題然而我懷疑LiDAR光學雷達能否自主而不受外界因素的影

因此，如果您支持LiDAR，請觀看此視頻。

光學雷達團隊：觀看此視頻。 它會有很多信息！

https://youtu.be/ibIzthKKZWY

如果您是支持特斯拉小組成員，請觀看此視頻

攝像團隊：觀看此視頻https://youtu.be/HM23sjhtk4Q

除了Tesla 。（Waymo，Nio，Uber，Ford，Toyota），幾乎每個人都在使用LiDAR。 你們可以看看埃隆·馬斯克（Elon Musk）在特斯拉團隊視頻中所說的話，然後自己判斷。

特斯拉將很快發布其FSD Beta 9.0。  （FSD =完全自動駕駛）。 您可以觀察特斯拉汽車在擁擠的城市情況下如何自行駕駛（目前有駕駛員監督）。

無論哪種情況，無論您是哪個團隊中，兩種技術的混合使用都可能是實現5級自治的最明智的解決方案。 在此之前，這樣的競爭遊戲仍在進行中！